PDF skill (Anthropic)

Quick answers

Who maintains this Claude Skill?

Anthropic, PBC (anthropics/skills repository).

What can it access?

It asks for: Run Python code that reads and writes PDF files (pypdf, pdfplumber, reportlab named in SKILL.md), Run command-line tools named in SKILL.md: pdftotext (poppler-utils), qpdf, pdftk if available, OCR scanned PDFs with pytesseract and pdf2image (SKILL.md notes a pip install is required), Run bundled scripts in scripts/ for form fields, bounding-box checks and PDF-to-image conversion (commands are given in forms.md, which SKILL.md tells Claude to read when filling a PDF form), Read and write local PDF files named in the task. Docs state: skills are "folders of instructions, scripts, and resources that Claude loads dynamically" (repository README), and the Help Center article says the feature requires code execution to be enabled. The pdf skill works on PDF files you provide or ask it to create. SKILL.md and forms.md do not mention sending file content to a third-party service, and the eight scripts in scripts/ contain no HTTP or socket calls (checked 2026-10-03). Where the data goes otherwise depends on the Claude product you use. (SKILL.md, forms.md and repository files, 2026-10-03). We do not label anything safe or unsafe; read the official sources before you install.

What licence does it use?

Proprietary, source-available (see limitations). Check the terms if you plan to use it commercially.

What are its limitations?

Licence detail: Proprietary, source-available (not open source). SKILL.md frontmatter: "Proprietary. LICENSE.txt has complete terms". LICENSE.txt begins "© 2025 Anthropic, PBC. All rights reserved." and says use is governed by your agreement with Anthropic regarding use of its services; if no separate agreement exists, by the Consumer Terms of Service or Commercial Terms of Service, as applicable. Under "ADDITIONAL RESTRICTIONS" it says users may not: extract the materials from the Services or retain copies outside the Services; reproduce or copy them, except for temporary copies created automatically during authorized use of the Services; create derivative works; distribute, sublicense, or transfer them to any third party; make, offer to sell, sell, or import any inventions embodied in them; or reverse engineer, decompile, or disassemble them. It adds that receipt, viewing, or possession does not convey any licence or right beyond those expressly granted. The repository README calls the docx, pdf, pptx and xlsx skills "source-available, not open source". Proprietary licence: LICENSE.txt lists additional restrictions on extracting or retaining copies outside the Services, copying, derivative works and distribution; read LICENSE.txt before any reuse. Repository README: provided "for demonstration and educational purposes only"; the behavior you get from Claude may differ from the repository version. OCR path needs extra Python packages (pytesseract, pdf2image) per SKILL.md; the Python and command-line tools it calls must exist in the environment. Requires code execution: the Help Center article says the feature requires code execution to be enabled, and the API guide says the code execution tool is required. The Claude Help Center article says: "Only install skills only from trusted sources." For a skill from a less-trusted source it says to review it before enabling, starting with the bundled files and paying particular attention to code dependencies and bundled resources like images or scripts.

When was it last released?

No tagged release shown (the repository releases list is empty on 2026-10-03). Latest commit on main: 2026-09-29. Verified Oct 3, 2026.